Agenda-setting

Rings Caused by Moon Collision

by Trevor Bailey

Until recently, researchers have been struggling to understand how Saturn’s rings formed and how old they are. However, a new study published this week claims that two of Saturn’s moons collided, creating the rings relatively recently.

Knowing that Saturn’s rings were affected by planetary dynamics, researchers set about looking into whether simulations of gravitational interactions in the past could reveal the origin of Saturn’s rings.

They found that the orbits of Saturn’s moons were disrupted about 100 million years ago. They suggest the Sun’s gravitational influence nudged one of Saturn’s moons into a collision with another, the debris from which formed Saturn’s rings.

Other scientists are skeptical of the study’s findings.

“There has been scientific debate in recent years on whether Saturn’s rings were formed by some kind of moon collision,” said Dr. Jonathan Hammig, who also researches astronomy and planetary science.

“The findings of this study contradict findings from previous research. Other studies have not found that the orbits of Saturn’s moons are different from what they used to be. There is still considerable disagreement on this topic within the scientific community.”

Hammig emphasizes that this study’s results are inconsistent with past research.

“This study challenges previous data we have suggesting that a moon collision did not create Saturn’s rings, so the results will really stir debate among scientists.”

Interactive Effect of Experimental Condition and Topic on Trust in the Commenting Scientist

Figure B.1: Interactive Effect of Condition and Topic on Trust in Commenting Scientist

Note: C = Agreement, T1 = Civil Disagreement, T2 = Uncivil Disagreement

Interaction Analyses and Results

There is no significant correlation between deference to scientific authority and conflict aversion (r = -0.03, p = .26).

We used stepwise OLS regressions to look at the main and moderating effects of partisanship, deference to scientific authority, and conflict aversion. The first step, OLS regressions including experimental condition, topic, partisanship, deference to scientific

Experimental Condition

authority, and conflict aversion can be found in Tables B.1-B.3. The second step included interactive effects of experimental condition and the moderator under investigation. The moderating effects of deference to scientific authority can be found in Tables B.4-B.6.

The moderating effects of conflict aversion can be found in Tables B.7-B.9. Visualization of significant interaction effects can be found in Figures B.2 and B.3.

Table B.1: Main Effects of Deference to Scientific Authority and Conflict Aversion on Outcomes (1/3)

Table B.2: Main Effects of Deference to Scientific Authority and Conflict Aversion on Outcomes (2/3)

Table B.3: Main Effects of Deference to Scientific Authority and Conflict Aversion on Outcomes (3/3)

Table B.4: Interactive Effects of Manipulation and Deference to Scientific Authority on Outcomes (1/3)

Table B.5: Interactive Effects of Manipulation and Deference to Scientific Authority on Outcomes (2/3)

Table B.6: Interactive Effects of Manipulation and Deference to Scientific Authority on Outcomes (3/3)

Trust in Scientists Trust in Scientific

Methods Better or Worse Benefits v. Harms Utility for Daily

Life Utility for

Table B.7: Interactive Effects of Manipulation and Conflict Aversion on Outcomes (1/3)

Table B.8: Interactive Effects of Manipulation and Conflict Aversion on Outcomes (2/3)

Table B.9: Interactive Effects of Manipulation and Conflict Aversion on Outcomes (3/3)

Trust in Scientists

Trust in Scientific

Methods Better or Worse Benefits v. Harms

Utility for Daily

Figure B.2: Visualization of Significant Interactive Effects of Deference to Scientific Authority and Experimental Condition

Figure B.3: Visualization of Significant Interactive Effects of Conflict Aversion and Experimental Condition

APPENDIX C

Pre-test Descriptive Information

Pre-test consensus estimate was measured by asking participants, “To the best of your knowledge, what percent (%) of climate scientists say that human activity is causing climate change? (0 - 100)” (M = 67.86, SD = 25.05).

Pre-test belief in climate change was measured by asking respondents, “How strongly do you believe that climate change is or is not happening?” Respondents reported their beliefs on a 5-point scale from “I strongly believe that climate change IS NOT happening” (1) to “I strongly believe that climate change IS happening” (5) (M = 4.10, SD = 1.17).

Pre-test belief in human causation was measured by asking respondents,

“Assuming climate change IS happening, to what extent do you think climate change is human-induced as opposed to a result of Earth’s natural changes?” Participants reported their beliefs on a 5-point scale from “Climate change is completely caused by natural changes” (1) to “Climate change is completely caused by human activity” (5) (M = 3.35, SD = 1.13).

Pre-test worry was measured by asking participants, “How worried are you about climate change?” Responses were captured on a 5-point scale from “Not at all worried”

(1) to “Extremely worried” (5). We also asked, “How concerned are you about climate

“Extremely concerned” (5). These measures of worry and concern about climate change were averaged to create a measure of pre-test worry and concern (M = 3.31, SD = 1.34, r

= .92, p = .00).

Pre-test support for public action was measured with the question, “What do you think of peoples' efforts to address climate change? Do you think that people should be putting more, less, or about the same amount of effort toward addressing climate change?” Participants responded on a 5-point scale from “People should put MUCH LESS effort toward addressing climate change” (1) to “People should put MUCH MORE effort toward addressing climate change” (5) (M = 4.09, SD = 1.16).

Finally, we asked participants about their pre-test support for government action with the question, “How strongly do you support or oppose government action to address climate change?” Responses were recorded on a 7-point scale from “Strongly support”

(1) to “Strongly oppose (7) (M = 5.32, SD = 1.71).

ANOVA Results

To look at condition differences with post-hoc Tukey tests, we first ran a series of ANOVA models with experimental condition and political ideology as predictors, along with gender, age, education, and concern about COVID-19. The omnibus F-tests for experimental condition and political ideology are reported below.

Consensus estimates were significantly associated with both experimental condition F(3, 1979) = 111.06, p = .00, hp2 = .144 and political ideology F(1, 1979) = 64.6, p = .00, hp2 = .031.

Belief in global warming was significantly associated with both experimental

condition F(3, 1986) = 3.45, p = .02, hp2 = .005 and political ideology F(1, 1986) = 305.35, p = .00, hp2 = .133.

Belief in human causation was significantly associated with both experimental condition F(3, 1986) = 5.64, p = .00, hp2 = .008 and political ideology F(1, 1986) = 247.59, p = .00, hp2 = .111.

Worry and concern about climate change was not associated with experimental condition F(3, 1986) = 2.33, p = .07. However, it was significantly associated with political ideology F(1, 1986) = 530.10, p = .00, hp2 = .211.

Support for public action on climate change was not associated with experimental condition F(3, 1985) = 1.41, p = .24. However, it was significantly associated with political ideology F(1, 1985) = 341.63, p = .00, hp2 = .147.

Support for government action on climate change was not associated with experimental condition F(3, 1985) = 2.01, p = .11. However, it was significantly associated with political ideology F(1, 1985) = 231.80, p = .00, hp2 = .105.

Coding Open-Text Responses

I coded two open-text questions for this study. The first asked participants what they thought informed their views on climate change. The second asked participants what they thought the survey was about.

To code responses to these items, I developed very basic dictionaries to capture language of interest. I did this by first looking at a random draw of 200 responses and making note of the language respondents used to express the quantities of interest. I then created simple dictionaries from this language. I then ran the dictionaries over the text

using approximate matching (agrepl in R). Following this, I human coded a random draw of 100 responses from each question and compared them to the dictionary coding. In both cases, the correlation between the human coding and dictionary coding was quite high (.92 for both, p < .001).

What influences your attitudes about climate change?

For the question about what influences one’s climate attitudes, I was specifically interested in how often scientific agreement was mentioned. It became clear from looking at the initial 200 responses that respondents did not often refer explicitly to consensus information (e.g., “if 97% of scientists say humans cause climate change, then i 100%

believe it”), but more frequently cited “scientists” as influences on their attitudes (e.g.,

“what the experts have to say about it” or “Scientists, health issues, ice Melting, sea rising”). I therefore coded any mentions of scientists, experts, or scientific agreement as indicating that the opinions of many scientists informed one’s climate attitudes (see dictionary in Table C.1). However, I did not include other mentions of scientific research, studies, or evidence in this coding. Though these are of course related to scientific

agreement, these responses did not refer explicitly to the opinions of experts, like was summarized in the consensus message, but could instead refer to models, individual studies, or other scientific facts.

In some cases, individuals talked about disagreeing with scientists in their responses. For example, one respondent wrote, “Common sense and the realization that most of these so-called scientists have agendas. Do you think they'll get any funding if they say, \"Everything is fine -- nothing to see here folks.\"? (sic)”. The dictionary results were not in any way corrected to deal with this type of negation. This appeared to be the

main source of inconsistencies between the human coding and the dictionary coding.

However, given that the correlation between human coding and the dictionary results remained high even without correcting dictionary results for these cases, I did not pursue any kind of correction to the dictionary results to account for people disagreeing with experts in their responses.

Though not quantitatively coded, other common influences on climate attitudes included personal experience or witnessed changes in weather (e.g., “Watching the devastating weather conditions around the world”); news, documentaries, and other media (e.g., I watch a lot of documentaries on nature and the natural world, including geology, and I have seen quite a bit of good documentaries showing the problems going on with climate change right now, especially at the polar caps.”); and environmental organizations and related appeals (e.g., “the polar bears are starving”). Others used this opportunity to express disagreement (e.g., “I am part of the 3% that believe it is a bunch of crap,” or “The 97% figure is a gross exaggeration, put forward by leftist to advance their anti-capitalist agenda”), conspiracy beliefs (e.g., “It's a front for the 1% agenda to form NWO”), or the belief that climatic changes are fated in some way (“Listen, God has control of the climate. It has nothing to do with man. If God wants to change the climate, then he will change the climate. It's in God's hand”). A few responses were humorous (e.g., “solid waste management and oil industry”).

What is this survey about?

In responses to the question about what this survey was about, I was particularly interested in whether people identified that the survey was about climate change, in spite of the distractor questions. Inspection of an initial 200 responses revealed that in most

responses that correctly identified that the survey was about climate change, participants included the phrases “climate change” or “global warming” (e.g., “various subjects, but more about climate change opinion”). In some cases, the responses were quite precise as to the intention of the study (e.g., “Seeing if these statements will change my opinion on climate change”). The dictionary for coding this item were therefore limited to the phrases “climate change” and “global warming” (see Table C.1). References to “the environment” were not included, as this sometimes appeared alongside mention of

climate change and other times along mentions of travel destinations (distractor questions included questions and images about travel).

There was evidence of reactance in responses to this question. For example, one respondent wrote that this survey was, “Ostensibly about climate change. In reality it is about a false statement that is used to manipulate opinions.” Others wrote about being forced or manipulated to hold a view on climate change (e.g., “Media trying to force me into thinking the climate change is all caused by humans” or “convincing me to believe your opinion that climate change is real”). Others inferred that politics was a central question (e.g., “I think and HOPE it's to show old republicans that anyone with a brain knows that climate change is real”).

Though not quantitatively coded, others identified the topic of the survey as related to travel and social media, which were the substance of the distractor tasks. Others took the opportunity to thank us for providing the survey. A few responses were

humorous (e.g., “Foolish chat-bots...Climate change!!!”).

Table C.1: Dictionaries for Open-Text Responses

Salience “scientists”, “experts”, “consensus”, “percent”, “97%”*

Demand Characteristics “climate change”, “global warming”

Note: I used approximate grep searches (agrepl) to approximately match keywords, thus allowing me to count misspelled keywords with the dictionary. *the 97% keyword was identified only using exact matching, because respondents occasionally referred to other percentages for unrelated reasons.